Page 272 - Advances in Renewable Energies and Power Technologies

6. Fault Detection Procedures 245

it has been mentioned before, the accuracy of the simulation results depends primar-
ily on the models and the parameter extraction techniques used.
Modeling and simulation of PV systems can be achieved by using Pspice
[69]. Pspice is a circuit simulation and analysis tool for analog and mixed-
signal circuits integrated in the OrCAD package developed by Cadence. This
software tool has demonstrated to be very useful in the simulation of PV systems
[46,70e72].
LabVIEW System Design Software is a sophisticated tool widely applied in in-
struments control, embedded monitoring, and control systems as well as in data
acquisition and signal processing [73]. It has been extensively used in monitoring
applications of PV systems [74e77] and also in simulation of PV systems
3
[78,79]. ANFIS was also used as a base for modeling and simulation of PV systems
[80]. However, MATLAB is probably the most used software in this field
[34,41e44,81e86]. MATLAB is a powerful technical computing environment
that can be complemented by a wide set of associated toolboxes offered by Math-
works [87]. It allows the modeling and simulation of PV systems and components
as well as monitoring PV plants. Moreover, it can also be combined with the Simu-
link interface, which is a friendly modular graphical environment of simulation,
resulting in a very powerful modeling and simulation platform [68,88].

6. FAULT DETECTION PROCEDURES
Two of the most used automatic supervision and fault detection procedures are
described in this section. The first one is focused in the analysis of the losses present
in the PV system, while the second one is based on the evaluation of current and
voltage indicators to detect faults and also identify the most probable fault present
in the system.

6.1 AUTOMATIC SUPERVISION AND DIAGNOSIS BASED ON POWER
LOSSES ANALYSIS
As it has been discussed in Section 2, the supervision of the PV system can be based
on the continuous check of normalized total inherent losses, L, described in Eq.
(7.5). For this purpose, theoretical boundaries based on the standard deviation, d, ob-
tained from a statistical study of simulated losses, L_sim, given in case of clear sky
conditions, must be defined.
The simulation of the PV array, having the PV module parameters, the configu-
ration of the PVarray, and the measured irradiance and module temperature as input
data, gives the expected values of the yields and L_sim as the result. On the other
hand, the supervision algorithm evaluates the evolution of the real yields from

3
Artificial neuro-fuzzy inference systems.